Sparse labeling of experimental tissue has been a powerful approach for cell biological analysis from the time of Ramon y Cajal, who used Golgi staining to describe in great detail the structure of neuronal tissue. To this date such labeling i critical for the analysis of neurons in whole tissue as the density and complexity of axonal and dendritic processes make tracing individual cells virtually impossible in ubiquitously labeled samples. Sparse labeling techniques also play critical roles in examining the behavior of cells during migration, in determining lineage relationships of cells, and assessing the cell autonomy of gene function. A variety of approaches have been used to sparsely label samples both in fixed and living tissue including Golgi staining, single cell dye injections, transient expression f DNA constructs, viral infection and cell transplantation. In vertebrate systems, several transgenic approaches to sparsely label cells have also been developed including use of promoters that display extensive variegation in expression, drug inducible activation of promoters and stimulation of mitotic interchromosomal recombination. However, none of these approaches for obtaining sparse labeling are easily applied for large-scale analysis in vertebrates (genetic screens or drugs screens). We propose to develop a robust, reliable method to create transgenic animals that will sparsely label a cell population and be readily reproduced from generation to generation. Conceptually, the idea is to use a newly described CRISPR/Cas9 site-specific nuclease to drive non-homologous end joining DNA repair in select cells to modify a non-expressed GFP reporter into one that expresses. If successful the approach should be widely applicable for the analysis of neuronal structure (and other cell types also) in mouse or zebrafish models of human neuronal diseases.